We uncovered ads success factors using behavioral science and machine learning.
The Problem
Our client routinely posted product-related banners on its homepage, but some of them had low engagement. The client wanted to know if they should use more cats or dogs in the ad to induce higher engagement. We believed that the issue was far more complex than that, as dogs or cats are displayed in banners to tell stories and resonate with the audience’s need or desire. It’s the story that ultimately matters.
The Audit
We developed a coding mechanism that identified the themes (e.g., recycling), emotions (e.g., pride), motivations (e.g., personal growth) that each of the 500+ ads was evoking, as well as the objectives that each ad included (e.g., dogs, cats). We coded each ad with 20+ identifiers and analyzed the data with machine learning in order to understand which elements of an ad led to higher engagement.
The Result
We found that ads that generate a sense of belongingness or empowerment, among a few other factors, had the highest engagement. Also, merely evoking positive emotions was not as favorable as one would think; having a dog or a cat ranked very low with respect to its impact on engagement.